Galaxy S23 Leak ChatGPT and Bing Father of Big Bang Theory 'The Last of Us' Recap Manage Seasonal Depression Tax Refunds and Identity Theft Siri's Hidden Talents Best Smart Thermostats
Want CNET to notify you of price drops and the latest stories?
No, thank you

Researchers work to develop prescription playlists

By studying a song's rhythm, melodic range, lyrics, and pitch, a team of scientists at Glasgow Caledonian University hopes to help regulate people's moods.


Prescription playlists may sound a bit out there, but even without all the findings in recent years that music can be good for the heart, the brain, and even the immune system, it should come as little surprise that it also affects mood.

Exactly how it does so is where things start to get interesting.

"The impact of a piece of music on a person goes so much further than thinking that a fast tempo can lift a mood and a slow one can bring it down," says audio engineering specialist Don Knox, who is leading a team at Glasgow Caledonian University in the UK that is studying songs' rhythms, melodic ranges, lyrics, pitches, and more to learn how music can regulate people's moods.

The three-year project, called Emotion Classification in Contemporary Music, is due to be complete in October 2010, and the team is analyzing not only the structures of songs but also the lyrics and even subjective factors such as whether certain songs are associated with happy or sad events, etc.

The team of researchers analyzed these factors in many types of songs, and they then play songs that volunteers say they have not heard before. On a graph, one axis measures the activity and intensity of the piece, and on the other each volunteer measures the feelings (from positive to negative) said piece arouses.

The researchers then look for patterns, such as what percentage of volunteers responds positively to a faster tempo and brighter timbre, or what happens when a major key suddenly jumps into diminished minor.

Knox says that the end goal is to develop a mathematical model that explains in great detail music's ability to communicate emotions, so that in a matter of years computer programs might identify which pieces of music is likely to influence a person's mood in a given direction.

So instead of clicking on a movie via Netflix or a coffee grinder via Amazon and being told what people who liked that also like, our music players might someday show little emoticons for each song and recommend other songs to fit that mood.

There is plenty of room for error, of course, the most obvious being that not everyone responds to music identically. For instance, I am one of those weird people who prefer to dance to The Cure, not cry. But if people can categorize their own music libraries, the computer algorithm might be able to use this historical association between certain songs and certain emotions in addition to the science and better cater to individual preferences.

"By making it possible to search for music and organize collections according to emotional content, such programs could fundamentally change the way we interact with music," Knox says. "Some online music stores already tag music according to whether a piece is 'happy' or 'sad.' Our project is refining this approach and giving it a firm scientific foundation, unlocking all kinds of possibilities and opportunities as a result."

Revealing our emotional responses to different songs runs the risk of being reductive, but the researchers predict that prescription playlists are not too far in the future, and that they could help motivate people when exercising, focus when studying, and even cope with physical pain.

I predict the algorithm becoming a widely-used feature on social networking and dating sites, too. I can see mine now: Fever Ray = smiley face with eyes all aglitter. Miley Cyrus = finger in mouth. That my husband and I have emotional responses in common is, after all, what brought us together in the first place.